Exploring machine learning techniques for fault localization

Ascari, Luciano C.; Araki, Lucilia Yoshie; Pozo, Aurora; Vergílio, Silvia Regina

doi:10.1109/latw.2009.4813783

“…These individual rankings are then consolidated to form a final statement ranking which can then be examined to locate the faults. There are other studies in this category such as those reported in [3].…”

Section: Machine Learning-based Techniquesmentioning

confidence: 95%

Software Fault Localization

Wong¹,

Debroy²

2010

Encyclopedia of Software Engineering

View full text Add to dashboard Cite

“…Wong et al (2009; also proposed two neural network-based fault localization techniques trained on program traces. Ascari et al (2009) investigate the use of SVMs in a similar setting.…”

Section: Probabilistic Debugging Methodsmentioning

confidence: 99%

Learning Tractable Probabilistic Models for Fault Localization

Nath

¹

,

Domingos

²

2016

AAAI

View full text Add to dashboard Cite

In recent years, several probabilistic techniques have been applied to various debugging problems. However, most existing probabilistic debugging systems use relatively simple statistical models, and fail to generalize across multiple programs. In this work, we propose Tractable Fault Localization Models (TFLMs) that can be learned from data, and probabilistically infer the location of the bug. While most previous statistical debugging methods generalize over many executions of a single program, TFLMs are trained on a corpus of previously seen buggy programs, and learn to identify recurring patterns of bugs. Widely-used fault localization techniques such as TARANTULA evaluate the suspiciousness of each line in isolation; in contrast, a TFLM defines a joint probability distribution over buggy indicator variables for each line. Joint distributions with rich dependency structure are often computationally intractable; TFLMs avoid this by exploiting recent developments in tractable probabilistic models (specifically, Relational SPNs). Further, TFLMs can incorporate additional sources of information, including coverage-based features such as TARANTULA. We evaluate the fault localization performance of TFLMs that include TARANTULA scores as features in the probabilistic model. Our study shows that the learned TFLMs isolate bugs more effectively than previous statistical methods or using TARANTULA directly.

show abstract

“…Wong et al (2009; also proposed two neural network-based fault localization techniques trained on program traces. Ascari et al (2009) investigate the use of SVMs in a similar setting.…”

Section: Probabilistic Debugging Methodsmentioning

confidence: 99%

Learning Tractable Probabilistic Models for Fault Localization

Nath¹,

Domingos²

2015

Preprint

0

View full text Add to dashboard Cite

In recent years, several probabilistic techniques have been applied to various debugging problems. However, most existing probabilistic debugging systems use relatively simple statistical models, and fail to generalize across multiple programs. In this work, we propose Tractable Fault Localization Models (TFLMs) that can be learned from data, and probabilistically infer the location of the bug. While most previous statistical debugging methods generalize over many executions of a single program, TFLMs are trained on a corpus of previously seen buggy programs, and learn to identify recurring patterns of bugs. Widely-used fault localization techniques such as TARANTULA evaluate the suspiciousness of each line in isolation; in contrast, a TFLM defines a joint probability distribution over buggy indicator variables for each line. Joint distributions with rich dependency structure are often computationally intractable; TFLMs avoid this by exploiting recent developments in tractable probabilistic models (specifically, Relational SPNs). Further, TFLMs can incorporate additional sources of information, including coverage-based features such as TARANTULA. We evaluate the fault localization performance of TFLMs that include TARANTULA scores as features in the probabilistic model. Our study shows that the learned TFLMs isolate bugs more effectively than previous statistical methods or using TARANTULA directly.

show abstract

Exploring machine learning techniques for fault localization

Cited by 31 publications

References 29 publications

Software Fault Localization

Software Fault Localization

Learning Tractable Probabilistic Models for Fault Localization

Learning Tractable Probabilistic Models for Fault Localization

Contact Info

Product

Resources

About